Growing more stem cells from one cord: ex vivo expansion and graft engineering

The first umbilical cord blood transplant, done in Paris in 1988 by Eliane Gluckman's team, treated a child with Fanconi anemia. It worked, and for the next decade cord blood transplants were mostly a pediatric story. The reason was not biology so much as arithmetic. A cord blood unit holds a small, fixed number of cells, and the bigger the patient, the more thinly those cells are spread.

This page explains that arithmetic, the workarounds transplant teams have used (two cords, engineered grafts) and the methods that grow more blood-forming cells outside the body, known as ex vivo expansion. Two expanded cord blood products have now reached regulatory approval, one in the US and one in the EU.

The arithmetic: cells per kilogram

Transplant doses are counted per kilogram of the recipient's body weight. Labs report two numbers: total nucleated cells (TNC), a rough count of all cells with a nucleus, and CD34-positive cells, a closer proxy for the stem and progenitor cells that rebuild blood.

A cord blood collection is small, usually somewhere around 50 to 150 milliliters. Transplant guidelines commonly ask for a minimum near 2.5 x 10^7 total nucleated cells per kilogram for blood cancers, and more for some inherited diseases.

Run the numbers on a hypothetical large unit holding 2 billion nucleated cells:

Many units are smaller than that, and some cells are lost during thawing and washing. So an adult can often find a well-matched unit that is simply too small.

Dose matters because it sets the speed of recovery. After the patient's own marrow has been wiped out by conditioning chemotherapy or radiation, there is a gap with almost no white blood cells. With cord blood, neutrophil recovery (the return of the main infection-fighting white cell) tends to be slow: in the control arm of the trial behind the first US-approved expanded product, the median was 22 days. Every day in that gap carries a risk of serious bacterial and fungal infection, and low doses also raise the risk that the graft never takes.

Two cords instead of one

In the early 2000s a University of Minnesota team reported giving adults two partially matched cord blood units at once, roughly doubling the dose. Double cord transplants became common for adults in the US.

A curious thing happens afterward. In most patients, one unit wins: within weeks, blood production comes almost entirely from one of the two cords, and the other fades. Why one dominates is still not fully clear.

The extra unit is not free. A randomized trial run by the US Blood and Marrow Transplant Clinical Trials Network, published in the New England Journal of Medicine in 2014, compared one versus two units in children and adolescents with blood cancers who had an adequately sized single unit. Two units did not improve survival at one year, and the double cord group had more graft-versus-host disease on some measures. For adults without a large enough single unit, double cords remained an option, but the trial showed that more cells from more donors is not automatically better.

Every additional unit also has to be found, matched and paid for. How cord blood tolerates partial HLA mismatch, and why that matters for patients of mixed or underrepresented ancestry, is covered in our piece on cord blood, genetic diversity and finding a match.

What "graft engineering" means

Graft engineering is any deliberate processing that changes what is in a stem cell graft before it is infused. It happens in hospital cell processing labs and, for manufactured products, in dedicated facilities. Common forms:

• Volume and red cell reduction. Cord banks routinely strip out most red cells and plasma before freezing, so units take less tank space and are safer to infuse.
• CD34 selection. Magnetic beads coated with an antibody against CD34 pull out the stem and progenitor cells, and most of everything else, including T cells, is discarded. This lowers the risk of graft-versus-host disease, at the price of slower immune recovery and more infections. It is used mainly with adult donor grafts, which start with far more cells than cord blood.
• T cell depletion. Instead of keeping only CD34 cells, labs can remove particular T cell types, such as the alpha-beta T cells that drive most GVHD, while keeping other immune cells that fight infection. This is used especially in half-matched (haploidentical) transplants for children.
• Expansion. The most ambitious version: growing more of the useful cells than the donor provided.

Our page on stem cell technology covers the basics of isolation, culture and cryopreservation that all of these rely on.

How an expanded cord graft is made, step by step

The details differ between methods, but the steps below follow the published description of the nicotinamide-based product approved by the FDA in 2023.

1. Choose the unit. The transplant team searches registries and public banks for a suitably matched cord blood unit, the same first step as an ordinary cord transplant. (Our page on cord blood banking explains where these units come from.)
2. Ship it frozen to the manufacturing site. Expansion is not done at the bedside. The unit travels in a liquid nitrogen dry shipper to a central facility.
3. Thaw and split. The cells are separated into two fractions. Magnetic selection pulls out cells carrying CD133, a marker found on early blood-forming cells.
4. Grow the selected fraction. The CD133-positive cells are cultured for about 21 days with four growth factors (stem cell factor, FLT3 ligand, thrombopoietin and interleukin 6) plus nicotinamide, a form of vitamin B3. Nicotinamide slows the cells' drift toward maturity, so more of them stay early progenitors as their numbers rise.
5. Hold the rest unchanged. The non-selected fraction, which includes the T cells and other mature immune cells, is kept frozen without culture.
6. Test before release. Each batch is checked for cell counts, viability, identity, sterility and other release criteria.
7. Infuse both. The patient receives conditioning on a schedule timed to the product's arrival, and both fractions are infused on transplant day.

The whole process takes weeks. That matters for a patient with fast-moving leukemia who cannot wait long.

Expansion methods that have reached patients

Growth factor cocktails alone

The earliest attempts, in the 1990s and 2000s, cultured cord cells with growth factors alone. Cell numbers rose, but mostly in more mature cells that did not shorten recovery much. These attempts showed that counting cells is not enough: what matters is keeping stem cell properties while numbers grow.

Notch ligand

Colleen Delaney and colleagues at the Fred Hutchinson Cancer Research Center reported in Nature Medicine in 2010 that culturing CD34 cord cells on an engineered form of the Notch ligand Delta1 produced cells that, given alongside an unmanipulated unit, led to faster neutrophil recovery in a small early study.

Co-culture with stromal cells

A team at MD Anderson Cancer Center grew cord blood cells together with mesenchymal stromal cells, mimicking the support cells of marrow, and reported in the New England Journal of Medicine in 2012 faster neutrophil engraftment than in historical controls.

Nicotinamide

The method behind Omisirge (omidubicel), described above, and the first expanded cord product approved by the FDA.

Aryl hydrocarbon receptor blockade (SR-1)

A 2010 Science paper identified StemRegenin 1, a small molecule that blocks the aryl hydrocarbon receptor and lets cord blood stem cells multiply in culture. A University of Minnesota phase 1/2 study, published in Cell Stem Cell in 2016, reported large increases in CD34 cells and quick engraftment.

UM171

A small molecule found at the Université de Montréal and reported in Science in 2014. Early clinical results, published in The Lancet Haematology in 2020, suggested expansion could let doctors pick smaller but better matched units. In August 2025 the European Commission granted conditional marketing authorization to a UM171-expanded cord blood product (Zemcelpro) for adults with blood cancers who need a transplant and have no other suitable donor.

What the 2023 FDA approval actually measured

On April 17, 2023, the FDA approved Omisirge for adults and children 12 and older with blood cancers who are planned for cord blood transplant after myeloablative conditioning. The stated purpose is to shorten the time to neutrophil recovery and lower the rate of infection.

The evidence came from a randomized trial of 125 patients comparing it with a standard cord blood transplant:

• Median time to neutrophil recovery: 12 days with the expanded product, 22 days with standard cord blood.
• Bacterial or fungal infections by day 100: 39 percent versus 60 percent.

Note what is missing. The approval was not based on longer survival or fewer relapses, and the trial was not designed to show them. The measured benefit is a shorter, safer gap after conditioning, which is clinically meaningful but narrower than "a better transplant". The product carries boxed warnings, as standard cord blood products do, about infusion reactions, graft-versus-host disease, engraftment syndrome and graft failure.

Limits and costs

Expansion solves the dose problem, not every problem.

• Time. Weeks of manufacture plus shipping and scheduling, and a plan for what happens if a batch fails release testing.
• Cost. A manufactured cell product adds substantial cost on top of the transplant itself, and coverage decisions vary by insurer and country.
• Fragile supply. One central manufacturing site and one small company mean a narrow pipeline. The developer of the US product went through a financial restructuring in 2024, a reminder that approval does not guarantee easy access.
• Competition. Half-matched family donor transplants, made safer by giving cyclophosphamide after the transplant, grew quickly through the 2010s, and cord blood transplant numbers fell in many centers as a result.
• Long-term questions. Faster early recovery is proven; whether expanded grafts sustain blood production for decades, as unmanipulated grafts can, is still being followed.

Expansion widens the set of usable units, but it does not change which diseases a transplant treats. That list, and why the "80 diseases" figures circulating online need context, is on our page on conditions treated with cord blood transplants. For the patients who gain most, those without a matched adult donor, the practical effect is that a well-matched unit too small to use on its own can now be considered.